Shrub Cover Models

Two models for estimating shrub cover percentage are available in Heureka.

Bilberry and lingonberry cover models by Bohlin

version 1.0 updated 25.4.2023

Model calculates the estimated bilberry and lingonberry shrub cover (%) based on the common stand attributes. Cover models are created based on the field data from Swedish NFI using general linear mixed effect models. Bilberry shrub cover is divided in two models, one for when site index is pine and one for when site index is spruce.

Model estimates should be used as an indication of the potential shrub cover, not as absolute measure.

Variables in Bilberry shrub cover models:

• When Site index for pine: Vegetation type class, temperature sum, soil moisture class, basal area, stand age, site index class, time since thinning
• When Site index for spruce: Vegetation type class, temperature sum, basal area, stand age, site index class, time since thinning

Variables used in Lingonberry shrub cover model:

• Vegetation type, temperature sum, basal area, basal area for pine, stand age, site index, time since thinning and time since clearcut.

Report including detailed description of the model will be published soon!

Models predicting the ground cover (%) of bilberry (Vaccinium myrtillus), cowberry (V. vitis-idaea) and heather (Calluna vulgaris) by Hedwall

version 1.0 updated 04.05.2023

Based on stand, soil and vegetation variables these models predict the percentage of the ground that is covered by either of the three dwarf-shrub species in summer.

The models are Generalized Linear Mixed Models, that accounts for zero-inflation, based on data from the Swedish National Forest Inventory (NFI). The models predict the cover based on the following explanatory variables:

• The proportion of the ground that was covered by the focal dwarf-shrub 10 years earlier
• The ratio of carbon and nitrogen in the top-soil (C/N)
• The stand age (years)
• Basal area (m2/ha) of trees
• The proportion of the basal area that is constituted by Norway spruce (Picea abies)
• The proportion of the basal area that is constituted by broadleaved tree species

The models therefore require a special database containing data for inventoried cover for respective species and the carbon:nitrogen-ratio in the humus layer. At present, such data only exists as a special database containing a subset of NFI plots, and their representative area is not rescaled if only the plots with necessary data is selected.

Report including detailed description of the model will be published soon!